Occhi sul Molo — Audio Chain

Percorso didattico completo dall'acquisizione WAV alla FFT e alle viste tempo-frequenza.

Indice

Tipo di demo

Generatorocchi_sul_molo_audio_chain 1.0.1
CategoriaAudio and Music
Sorgenteregistrazione stereo originale dell'autore, distribuita con il demo.

1. Obiettivo

Seguire una catena completa e tracciabile da un WAV originale ai Dataset numerici e alle rappresentazioni spettrali.

Workflow Occhi sul Molo
La Session fresca contiene la sorgente e la guida; una Session demo eseguita separata contiene la catena già completata.

2. Sorgente

WAV 48 kHz, stereo, 16-bit PCM, circa 284.4 s. Il SHA-256 viene registrato in metadata/audio_source.sha256.

3. Percorso consigliato

  1. Import WAV con WAV acquisition.
  2. Audio Decode → canonical PCM Dataset.
  3. FFT Spectrum su intervallo o blocco.
  4. Spectral Features su spectrum.table.
  5. Temporal Window Set da 1 s con step 0.5 s.
  6. Time-dependent FFT → immagine 2D, superficie 3D e confronto di finestre.

4. Cosa osservare

Sample rate, channel count, durata, provenance, spectrum contract, centroid, bandwidth, rolloff, flatness e cumulative spectral power. Il dominant bin di una singola finestra non coincide automaticamente con il “pitch della canzone”.

5. Nota sulle metriche

THD, SINAD ed ENOB su una FFT musicale sono diagnostiche della procedura numerica selezionata, non rating sensati della qualità di un brano polifonico.

Riferimento tecnico completo

Student guide

Occhi sul Molo: audio acquisition to spectral views

This Session contains the author’s original Occhi sul Molo WAV (48 kHz, stereo, 16-bit PCM; about 284.4 seconds). The author authorized its distribution for this GX-TXT demonstration. The WAV’s SHA-256 is recorded in metadata/audio_source.sha256.

Suggested workflow

  1. Import raw/Occhi_Sul_Molo_lavoro.wav using WAV acquisition. Preview shows a bounded binary summary and does not read the entire file into the GUI.
  2. Run Audio Decode to publish the canonical PCM Dataset. Check sample rate, channel count, duration and provenance.
  3. Run FFT Spectrum & Error Analysis on a chosen interval or fixed block. The resulting spectrum.table is an audio frequency spectrum; spectroscopy’s spectrum.scientific is a separate contract.
  4. Run Spectral Features on that FFT result. Compare dominant bin, spectral centroid, bandwidth, rolloff, flatness and cumulative spectral power. The dominant bin from one short window is not the song’s musical pitch.
  5. Create a 1-second Temporal Window Set stepped by 0.5 seconds, then run Time-dependent FFT. Inspect the 2D time-frequency image, 3D surface and a Plot Variant comparing early, middle and late windows.
  6. Use the experiment and Dataset Markdown editors to record assumptions, settings and conclusions. Rerender Plot Variants from cached scientific data, then export the Session as an offline HTML site.

The separately supplied executed faculty DEMO Session includes the real analyses, reports, rich notes and audited HTML. The generator creates a fresh teaching Session so the entire workflow can be repeated.

THD, SINAD and ENOB from a music FFT are diagnostics of the selected numeric procedure, not meaningful audio-quality ratings for a polyphonic song.

Generator README

Occhi sul Molo audio Session Generator 1.0.0

This portable GX-TXT Session Generator creates a new Session with the author’s original stereo WAV, checksum record and teaching guide. Import the WAV and run the analyses using GX-TXT; the separate executed DEMO Session contains the completed chain.

The package is self-contained and does not fetch external media.

Rights and provenance

Rights and provenance

The recording Occhi sul Molo is supplied by its author, who explicitly stated that they own the rights and that it may be distributed for this GX-TXT demonstration. This statement is the basis for including the exact WAV whose SHA-256 is 09d2235cb618fd65d834124f7c0af1fa023d6ef3b579de60b45e0d538b26cdcc. No general third-party license for the recording is asserted.

The generator script and written teaching materials are provided by Giux-Lab for this private GX-TXT release. The source music-lab Session was not modified to make this package.

Generator descriptor

; GX-TXT Session Generator API v1
[generator]
api_version=1
generator_id=occhi_sul_molo_audio_chain
name=Occhi sul Molo — complete audio chain
category=Audio and Music
description=Author-owned stereo recording for WAV acquisition, decoding, FFT, time-frequency and spectral-feature teaching.
version=1.0.1
generator_type=python
entry_point=generate.py
author=Giux-Lab
tags=audio;music;wav;fft;spectrogram;spectral-features;teaching
required_capabilities=Python 3;GX-TXT Session
suggested_session_name=occhi-sul-molo-audio-chain
expected_purpose=Reproduce a traceable audio workflow from an author-owned original WAV to numerical and graphical Datasets.
files_produced=raw/Occhi_Sul_Molo_lavoro.wav;metadata/audio_source.sha256;STUDENT_GUIDE.md